Multimedia over coax alliance (MoCA) technology for home networks is well-known and well-understood by someone skilled in the art, as shown and described in www.MoCAlliance.org. MoCA is commonly used to form a home network, which conveys Ethernet frames among different rooms in the home over existing coaxial cabling. MoCA's traffic model is known as multipoint-to-multipoint, because generally every MoCA device in the home can communicate directly with any other MoCA device in the home. One of the MoCA devices in the home is designated the Network Coordinator and becomes responsible for coordinating and scheduling all traffic on the MoCA home network. MoCA devices form burst transmissions that are launched on the coax cables as radio-frequency (RF) signal transmissions. These transmissions propagate over the coax medium to reach one or more or all of the other MoCA devices in the home network. The Network Coordinator schedules the time and spacing of individual burst transmissions so they do not destructively collide (e.g., overlap in time) at the intended receiver(s). This scheduled arrangement for sharing a communication channel on the coax medium for traffic from and to various endpoints is commonly known as time-division multiple access (TDMA).
Entropic's c.LINK® technology is a coax access system that has some elements similar to MoCA technology for Access Networks. The c.LINK access networks have been deployed in China to service multiple-dwelling units (MDU) (e.g., apartment buildings) over coax cable plants. These adaptations include:                a) locating the Network Coordinator at a headend site;        b) locating individual c.LINK devices as customer premise equipment (CPE);        c) scheduling traffic as time-division duplex (TDD);        d) scheduling downstream traffic from the headend-to-CPEs, aka point-to-multipoint;        e) scheduling upstream traffic from CPEs-to-headend, aka multipoint-to-point.        
Access Networks are typically deployed by operator/service providers (OSPs) to provide paid high-speed access to the internet and other OSP-provided services. This includes residential services (such as pay TV, telephony, and internet data), as well as business services for businesses seeking strict quality of service (QoS) service level agreement (SLA) contracts including low-latency, low-jitter, and guaranteed throughput. The OSP typically deploys equipment at some headend (e.g., located in an OSP's central office site, or near some residential neighborhood, or in the basement of an MDU), which communicates with each of one or more (typically a plurality of) customer premise equipments (CPEs) deployed at endpoint sites such as individual residences or businesses. The headend can transmit messages downstream (DS) over the access network to CPEs (aka point-to-multipoint), and the CPE endpoints can transmit messages upstream (US) to the headend in the opposite direction over the access network (aka multipoint-to-point). Access networks can be based on various media types, such as:                a) Fiber optic cabling;        b) Coaxial metallic cabling;        c) HFC hybrid fiber and coax cabling (e.g., as identified in http ://www.ieee802.org/3/epoc/public/mar12/schmitt_01_0312.pdf);        d) Other metallic cabling (e.g., twisted-pair copper subscriber loops);        e) Other hybrids, such as fiber and twisted-pair; and,        f) Wireless media, such over the air transmissions.        
The access network medium may contain a cascade of various active components (such as signal amplifiers), as well as lossy passive components (such as splitters or taps), deployed at various fixed locations in the network. Also, the distances, or medium pathlength, from the headend will generally vary for each CPE. These differences generally result in differing propagation times among the various branches in the network, as well as differing arrival amplitudes of transmitted signals. Consequently, the path between the headend and each CPE will vary, and the associated signal attenuation, or pathloss, will vary correspondingly. If the pathloss is relatively low, the CPE may be characterized as nearby the headend. If the pathloss is relatively high, the CPE may be characterized as distant from the headend.
Upstream (US) transmissions are formed and launched by CPEs, but are generally not continuous, so upstream traffic from a plurality of CPEs is typically coordinated by the headend in order to ensure that those non-continuous or burst transmissions from various CPEs do not collide (overlap in time) and that the headend receiver will observe an orderly sequential arrival of burst transmissions from different CPEs in a predictable order and at predictable times (within some tolerance of time-jitter). This approach is often called time-division multiple access (TDMA).
Some OSPs operate their access network such that upstream traffic and downstream traffic use different frequencies or wavelengths, enabling transmissions in both directions, simultaneously and independently (i.e., full duplex). This particular duplexing strategy is called frequency division duplex (FDD). The headend has exclusive use and access to the downstream frequencies, and the headend can coordinate/schedule use of the upstream frequencies independently from the downstream. FDD is relatively common on access networks today, even though FDD operation does incur overheads that reduce spectral efficiency, such as the spectral guard band imposed by inflexible diplex filters distributed throughout the access network cascade (e.g., to isolate the US and DS traffic from each other).
However, many OSPs desire a different mode of operation: time-division duplex (TDD), where a single RF spectral channel-width (or single optical wavelength) is being used, and alternating-in-time between upstream and downstream (half duplex). TDD's single half-duplex channel alternates between upstream (US) and downstream (DS) traffic, which implies the DS link would be unavailable during US traffic, and vice versa. OSPs wish to consider such a mode due to TDD's increased flexibility (compared to FDD) for adapting to the evolution of future US and DS traffic patterns.
One benefit of TDD is that the symmetry or asymmetry of the US and DS capacities is a relatively simple (and realtime) adjustment of the duty-cycle phasing of the TDD Cycle. Use of TDD in the Access Network enables a more flexible way for OSPs to easily, quickly and inexpensively adjust the relative throughput capacity of the upstream and downstream directions within a single spectral allocation. Whereas, FDD requires paired spectral allocations typically established by inflexible diplex filters distributed throughout the access network cascade. For a given total aggregate spectral allocation, TDD's single spectral allocation can be made as wide as the sum of FDD's paired allocations, enabling TDD's burst datarate capability in either direction being approximately double that of FDD in either direction (for symmetric US and DS FDD allocations). A more extensive discussion of TDD is available at the website located at http://www.ieee802.org/3/epoc/public/may12/barr_01_0512.pdf.
Some access networks currently deploy TDD technology, such as c.LINK Access, which is typically used to service MDUs in China. In addition, there are new access network technologies currently under development (e.g., in the IEEE 802.3bn EPoC Task Force, and in ITU-T G.fast) which are being designed to leverage the inherent flexibility of TDD.
TDD operation has certain overheads that reduce temporal efficiency. For example, it is necessary for the headend scheduler to avoid collisions in the TDD mode of operation by segregating the upstream traffic from downstream traffic with a time gap, such as an inter-phase gap (IPG) between TDD phases. An IPG may include time intervals for transmissions to complete their propagation from transmitter(s) to intended receiver(s), time for the medium to sufficiently quiesce (if necessary) after reception(s), and time for destination transceiver(s) to switch (if necessary) from receive to transmit mode. As shown in FIG. 1, the sequential combination of any two adjacent TDD phases (i.e., a single downstream phase 12 plus a single upstream phase 16) is called a TDD Cycle 10. There are IPGs 14 between phases. TDD Cycles 10 with duration on the order of 1 millisecond are commonly deployed in access networks employing TDD (although longer or shorter durations can be used). In the TDD mode of operation, it is common for each transmitter to prepend a preamble signal at the beginning of each of its transmissions. These preambles include reference signals which can be useful to facilitate receivers in detecting and acquiring the physical layer (PHY) parameters required to properly decode the transmission, such as gain, frequency-offset and timing information. The time intervals that preambles consume on the medium, are generally accounted as overhead for TDD. In the upstream, the PHY of each CPE typically starts its burst transmission with a preamble as shown in FIG. 2, where preamble 18 is transmitted first, followed by payload 20 (message-information or user-data carrying portion) of the transmission.
These upstream preambles facilitate the headend receiver to detect and acquire the PHY layer parameters which in general are unique to each individual CPE device. For example, the differing pathloss from each CPE to the headend generally results in differences in arrival amplitude at the headend from each CPE's upstream transmissions. These differences in arrival amplitude correspond in general to different SNRs as received at the headend. Anyone skilled in the art knows that the capacity for a channel to carry information is closely related to this received SNR and is discussed in the website located at http://en.wikipedia.org/wiki/Channel_capacity.
When a CPE is nearby the headend, the pathloss may be low, the arrival amplitude may be high, and the SNR may be high, yielding a higher channel capacity to or from that particular CPE. Conversely, when a CPE is distant from the headend, the pathloss may be high, the arrival amplitude may be low, and the SNR may be low, yielding a lower channel capacity to or from that particular CPE. The headend, being aware of these differences in reception, generally schedules each CPE to transmit its payload information using a modulation profile that corresponds to the particular SNR that can be received from that particular CPE. For example, the headend might schedule a higher modulation profile for nearby CPEs, and a lower modulation profile for distant CPEs.
Modulation Profile (MP) generally refers to various combinations of modulation density with forward error correction, or MCS modulation and coding scheme, as discussed in the website located at http://en.wikipedia.org/wiki/Modulation_and_coding_scheme. Different Modulation Profiles are generally chosen to adapt communication signal transmissions to the particular conditions experienced on the communications channel. A high modulation profile generally corresponds to a relatively high-density modulation (e.g., 1024-QAM, being higher than 256-QAM), and/or a relatively high coding rate for forward error correction. A lower modulation profile corresponds to a relatively low-density modulation (e.g., 256-QAM, being lower than 1024-QAM), and/or a relatively lower coding rate for forward error correction. A high modulation profile carries more information bits per second (or per symbol) than a lower modulation profile. However, high modulation profiles are more difficult to receive and decode (i.e., low receptivity), requiring better channel conditions (e.g., higher SNR) than would otherwise be required with a lower modulation profile that is easier to receive and decode (i.e., having higher receptivity).
Summarizing, for distant CPEs whose channel conditions are insufficient, the headend must resort to transmitting lower modulation profiles and suffer the lower bits-per-second information rate. It would be better for the OSP if the headend could transmit high modulation profiles whenever possible, enabling nearby CPEs whose channel conditions are sufficiently good to realize the higher bits-per-second information rate. However, preambles represent the overhead associated with the headend changing modulation profiles, and these overheads work against the benefits that otherwise could be realized by using different modulation profiles.